@Vincent
My 040/25 was drastically quicker than my 030/30 (over double) so I don't think a few extra MHz will make the 030/50 faster than an 040.
Using SysInfo (SysSpeed refused to work with my 030) my 030/30 was averaging 7.2x mips, the 040/25 was 18.xx mips.
Yet again, MIPS is not meaningful when comparing different CPU models.
You can't draw such conclusions from simple MIPS values.
@LinchpiN
In my opinion 030@50 is very nice since it doesn't require any active cooling (like some 040 accelerators). IMO if you need faster system than 030@50, skip 040 altogether and get 060 directly.
